| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Department of Plant and Soil Sciences, 368 Ag. Hall, Oklahoma State University, Stillwater, OK 74078, USA
b International Maize and Wheat Improvement Center (CIMMYT), Km. 45, Carretera Mexico, El Batan, Texcoco, Mexico
c Department of Primary Industries (DPI), Private Bag 260, Horsham, Victoria, Postcode: 3401, DX Number: 216515, Australia
d Department of Biosystems and Agricultural Engineering, Oklahoma State University, Stillwater, OK 74078, USA
* Corresponding author (aklatt{at}mail.pss.okstate.edu)
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: GHIST, global historic trials • GNDVI, green normalized difference vegetation index • LAI, leaf area index • NIR, near infrared radiation • NWI-1, normalized water index-1 • NWI-2, normalized water index-2 • PRI, photochemical reflectance index • RLs1, random lines-1 • RLs2, random lines-2 • RNDVI, red normalized difference vegetation index • SR, simple ratio • SRI, spectral reflectance indices • VIS, visible wavelength • WI, water index
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
The use of morphological and physiological selection criteria to differentiate grain yield is an indirect breeding approach. Physiological tools have had limited utility in plant breeding programs (Jackson et al., 1996), partly because of the time-consuming evaluation methods and the lack of association with yield (Loss and Siddique, 1994; Richards, 1996). Nonetheless, canopy temperature, which can be sensed remotely using infrared thermometry, has been shown to be well associated with the yield of wheat cultivars (Reynolds et al., 1994; Fischer et al., 1998), as well with the yield of recombinant inbred lines and advanced breeding materials (Reynolds et al., 1998; 1999) in irrigated, high radiation environments.
More recent studies suggest that spectral reflectance is another promising remote sensing technique for screening genotypes (Araus, 1996; Araus et al., 2001). Canopy light reflectance properties mainly based on the absorption of light at a specific wavelength are associated with specific plant characteristics. The spectral reflectance in the visible (VIS) wavelengths (400–700 nm) depends on the absorption of light by leaf chlorophyll and associated pigments such as carotenoids and anthocyanins. The reflectance in the VIS is low because of the high absorption of light energy by pigments. The reflectance of the NIR wavelengths (700–1300 nm) is high because of the multiple scattering of light by different leaf tissues (Knipling, 1970). Spectral reflectance indices were developed on the basis of simple mathematical formulae such as ratios or differences between the reflectance at given wavelengths (Araus et al., 2001). Simple ratio (SR = NIR/VIS) and normalized difference vegetation index [NDVI = (NIR – VIS)/(NIR + VIS)] were the first SRI developed, combining information from the VIS and NIR wavelengths. These indices were used to predict different vegetation parameters, such as green biomass and green leaf area index (LAI) (Tucker and Sellers, 1986). Spectral reflectance indices have also been developed based only on VIS, including the photochemical reflectance index [PRI = (R531 – R570)/(R531 + R570)] used to assess radiation use efficiency by the plants (Peñuelas et al., 1995), and also only on NIR, such as the water index (WI = R900/R970) to assess water status of the canopy (Peñuelas et al., 1993). Spectral reflectance indices have been widely reported by different authors to assess different physiological conditions of the canopy such as total dry matter, LAI, photosynthetic capacity (Sellers, 1987), as well as green LAI and fraction of photosynthetically active radiation absorption (Wiegand and Richardson, 1990; Baret and Guyot, 1991; Wiegand et al., 1991). Spectral reflectance indices have also proven to be useful in the assessment of early biomass and vigor of different wheat genotypes (Elliott and Regan, 1993; Bellairs et al., 1996), water status in gerbera (Gerbera jamesonii) and barley (Hordeum vulgare L.) (Peñuelas et al., 1993, 1997), and different pigment concentrations in the leaves of soybean [Glycine max (L.) Merr.] (Chappelle et al., 1992). The potential for using SRI to predict in-season grain yield have also been reported in wheat (Raun et al., 2001) and in maize (Zea mays L.) (Osborne et al., 2002) under water-stressed environments.
Attempts have been made to evaluate the potential use of SRI in plant breeding to differentiate genotypes for yield under well-watered and/or moisture-stressed conditions in wheat (Hatfield, 1981; Ball and Konzak, 1993; Aparicio et al., 2000; Royo et al., 2003) and soybean (Ma et al., 2001). The studies under moisture stressed conditions showed the potential of using spectral indices under such conditions, but under well-watered conditions the association between yield and existing indices (NDVI and SR) were weak.
The goal of this study was to evaluate a broader range of SRI as potential screening tools in irrigated, high yielding environments. Specific objectives of the present study were to (i) evaluate the correlation of existing spectral indices with yield and agronomic traits of bread wheat genotypes under near optimum nitrogen and irrigation levels, (ii) derive new improved SRI that distinguish among high yielding genotypes better than preexisting indices, and (iii) to determine the best growth stage to apply the spectral reflectance tool.
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
The seeding rate for each experiment was 78 kg ha–1 and the experiments were planted in the last week of November. Nitrogen and phosphorus were applied to the plots at rates of 200 kg ha–1 and 26 kg ha–1, respectively. During the first 2 yr, 150 kg N and all of the P were applied during land preparation, and 50 kg N was applied in the second week of January coinciding with the first node growth stage and the second supplementary irrigation. In the third year, the same procedure was followed, but N was supplied at half the dosage during planting and half with the second supplementary irrigation. A total of five supplementary irrigations were applied in the first and third years, but in the second year only four supplementary irrigations were given. Folicur 250EW (25% tebuconazole) was applied twice in every crop cycle, in the second week of February (early booting) and in the second week of March (just after flowering) at the rate of 0.5 L ha–1 to protect the experimental materials from prevalent leaf rust (caused by Puccinia triticina Eriks).
Harvested area for grain yield was 4.8 m2 (1.6 by 3 m). Before harvesting, 100 tillers with spikes were cut at the ground level to estimate the various yield components. The collected 100 tillers were oven dried at 75°C for 48 h. The weight of the oven-dried 100 tillers was measured, and then the tillers were threshed to calculate the harvest index. Two hundred grains were randomly collected from the harvested plots to estimate 1000-grain weight. Harvested grain yield was converted to grain yield in megagrams per hectare (Mg ha–1). From harvest index, grain weight of 100 tillers, 1000-grain weight, and grain yield per unit area, other yield components, including spikes per square meter, grains per spike, and biomass at maturity were calculated.
Experimental Materials
Experiment 1: In the first experiment, we used 15 worldwide-adapted spring bread wheat genotypes developed by the wheat-breeding program of CIMMYT. The genotypes represent the historical success achieved by the breeding program at CIMMYT. While all are high yielding in distinct regions around the world, they vary widely in morphological traits and in parentage. The genotypes were planted in a 5 by 3 
-lattice design with two replications. In this paper we will refer to this experiment as "GHIST," since it studies a global historical set of commercial genotypes.
Experiment 2: This experiment had 25 genotypes, comprising 23 random F3–derived lines and the two parents Sonalika and Attila. Approximately 1000 F2 seeds were planted and the population was harvested in bulk. Next year approximately 500 F3 seeds were space planted and random individual F3 plants were selected. The selected F3 plants were harvested separately and F3:4 families were planted in separate small plots. The F3:4 small plots were harvested separately (F5 seed) and planted in individual F3:5 yield plots that were harvested separately. The process was continued for two more generations to produce F3:7 families. The experiment was planted in a 5 by 5 -lattice design with two replications. In this paper we will refer to this experiment as "RLs1."
Experiment 3: We used 36 genotypes, comprising 34 random F3–derived lines and their two parents (Bacanora 88 and Cndo/R143//Ente/Mexi_2/3/Ae. sq.(Taus)/4/Weaver). The random lines were developed in the same procedure as described for RLs1. The experiment was planted in a 6 by 6 -lattice design with two replications. In this paper we will refer to this experiment as "RLs2."
Radiometric Measurements
The spectral reflectance measurements were taken by a portable narrow-bandwidth Spectroradiometer (Model FieldSpec UV/VNIR, Analytical Spectral Devices, Boulder, CO) with a 25° field of view. This instrument can detect reflected light from the canopy ranging from 350 to 1100 nm. Therefore, it covers VIS and NIR. It gives 512 continuous bands with a sampling interval of 1.43 nm. The spectroradiometer was connected to a computer, which stored the individual scans for subsequent processing. Each reflectance measurement was the average of 10 scans (which was programmed and calculated by the software used to operate the spectroradiometer) and the scanning area was approximately 18.94 cm2. The sensor was mounted with the help of a pistol grip 40 to 50 cm above the canopy facing the center of the bed. The spectroradiometer was recalibrated against a white reference plate (BaSO4) every 10 plots. The reflectance measurements were taken between 1030 to 1400 h under sunny conditions, and reflectance measurements were taken from four different places within each plot. The mean of the four readings was used to calculate spectral indices of each individual plot for statistical analysis. The average time required for the completion of the reflectance measurement from four different areas within the plot was approximately 40 to 45 s plot–1.
The spectral reflectance measurements were taken at booting (Zadoks' stage between 39 and 47), heading (Zadoks' stage between 55 and 69), and grain filling (Zadoks' stage between 75 and 83) in all experiments (Zadoks et al., 1974), except for RLs1 and RLs2 in the year 2001–2002, where the reflectance measurements were taken only at the heading stage.
Calculation and Selection of Indices
Initially, different ratios and normalized indices were calculated based on a combination of visible and near-infrared wavelengths. From the combinations tested, two indices were selected for presentation in this paper. Those two indices combined information from 850, 900, and 970 nm. The 970 nm has been reported as a weak water absorption band (Peñuelas et al., 1993), and the other two bands (850 and 900 nm) were used as reference bands. We have referred to these two indices as normalized water index-1 (NWI-1) and normalized water index-2 (NWI-2). Five other reference indices, including the most widely used NDVI and SR, were calculated (described below) and compared with the two new indices.
The notation Ri was used to indicate the reflectance of light at a wavelength of i nm. The different SRI calculated were: PRI = (R531 – R570)/(R531 + R570) (Peñuelas et al., 1995), which is an indicator of radiation use efficiency by the plants; WI = R970/R900 (Peñuelas et al., 1993), which indicates canopy water status; RNDVI = (R780 – R670)/(R780 + R670) (Raun et al., 2001), which indicates canopy photosynthetic area; GNDVI = (R780 – R550)/(R780 + R550) (Gitelson et al., 1996), which indicates canopy photosynthetic area; SR = R900/R680 (Aparicio et al., 2000), which is also an indicator of canopy photosynthetic active area. The above mentioned normalized water indices were calculated as follows: NWI-1 = (R970 – R900)/(R970 + R900), and NWI-2 = (R970 – R850)/(R970 + R850).
Statistical Analysis
Alpha-lattice analyses for grain yield and spectral indices were performed using PROC MIXED procedure of the SAS/STAT statistical package (SAS Institute, 2001). Data were analyzed in each individual growth stage within the same year and between the years. Combined analyses were performed across different growth stages and different years (years 2002–2003 and 2003–2004 for RLs1 and RLs2, and all 3 yr for GHIST). Mean squares of combined analysis were obtained by using PROC MIXED following type1 method. Pearson correlation coefficients were used to estimate the relationships of yield and yield components with different spectral indices and the relationships of indices at different growth stages within the same year and between years. Genetic correlations between traits were estimated using PROC MIXED, using a program following the method described by Singh and Chaudhary (1977).
|
RESULTS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
|
|
|
|
|
Association of Grain Yield with SRI
The correlation of grain yield with spectral indices at (i) different growth stages for each year, (ii) averaged across growth stages for each year (mean), and (iii) averaged over growth stages and years (overall mean) are presented in Table 6. The association between mean SRI (averaged over growth stages within an individual year) and the overall grain yield (mean of 3 yr) are presented in Table 7. We calculated 72 different ratios and normalized indices by using different visible and NIR wavelength combinations and also calculated 20 previously published SRI which were indicative of different physiological conditions of plants. Initially, we selected five out of 72 SRI that were calculated but later selected only two of them to present in this paper because of their high correlation with grain yield and consistent performance over years and different genetic backgrounds. Of the previously published SRI, we selected five to present in this paper. RNDVI, GNDVI, and SR have been the most widely used by different authors to study the physiological conditions of plants, and all five SRI (RNDVI, GNDVI, SR, WI and PRI) have been reported by various authors to differentiate genotypes for grain yield under water-stressed conditions in durum wheat (Triticum turgidum L. subsp. durum) and well-watered conditions in bread wheat and soybean.
|
|
At booting, RNDVI, GNDVI, SR. and PRI were better correlated with grain yield than the NIR-based indices. Near infrared radiation–based indices showed a very low level of association with grain yield at the booting stage, which was consistent across all the experiments and in all the years. In most cases, mean indices over three growth stages correlated better with grain yield than any single growth stage with a few exceptions. A similar trend has also been observed when the mean indices over two growth stages (heading and flowering) were used, and the correlation values were very close to the correlation values when the mean indices over three growth stages were used, with a few exceptions (data not shown). The correlations between the overall mean indices (across different growth stages and years) showed very strong correlations with the overall mean grain yield of genotypes (across different years) and the correlations values were higher than between SRI and grain yield within an individual year.
The three NIR-based indices showed a very strong association with overall mean grain yield (mean of 3 yr), when the mean indices over three growth stages within an individual year were correlated with overall mean grain yield. The levels of association were consistently higher than the association between other SRI and overall mean grain yield with a single exception (Table 7).
No single NIR-based index showed any definite superiority over the others. Nonetheless, when we consider all the correlations between these three indices and grain yield within year, between indices measured in an individual year and overall mean grain yield, and between overall mean indices and overall mean grain yield, NWI-2 gave either similar or marginally better correlations than the other two NIR-based indices (WI and NWI-1). For simplicity of the presentation in the regression model, we have used only NWI-2.
Attempts were made to determine the most suitable regression model to explain variability among the genotypes for grain yield across different years (Fig. 1,2 ). NWI-2 and RNDVI were compared in different regression models because of the performance of NWI-2, and because the RNDVI has been the most widely used SRI to study grain yield variations at the genotypic level, as well as in agronomic trials. In general, a simple linear model did not differ greatly from exponential and power regression models in explaining grain yield variation (except in GHIST). The exponential model explained 2 to 3% more of the variability for grain yields for the different indices in year 2003–2004. A simple linear model was equally applicable in explaining the phenological pattern of the relationship between grain yield and the indices (except for GHIST) (data not presented). The exponential model explained 2 to 3% more of the variability for RNDVI and NWI-2 when the readings were taken at heading and grainfilling stages in year 2003–2004. As the nonlinear model was not significantly better than the linear model, simple linear models are presented in Fig. 1 and 2.
|
Association of SRI with Agronomic Traits
The associations between SRI and two agronomic traits, grains per square meter and biomass at maturity, averaged over the 3 yr are presented in Table 8. The correlations between SRI and additional agronomic traits such as harvest index, spikes per square meter, grains per spike, and 1000-grain weight were also evaluated. In general, the associations between SRI and these four agronomic parameters were low and inconsistent (not shown).
|
|
DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Effect of Growth Stages
In general, values of PRI, RNDVI, GNDVI, and SR decreased from heading to grainfilling (Table 2). Aparicio et al. (2000) also reported a similar trend of decreasing values for spectral indices (PRI, RNDVI, and SR) with the advancement of growth stages in durum wheat under irrigated conditions. As RNDVI, GNDVI, and SR are indices that combine red, green, and NIR wavelengths, a reduction in LAI would decrease the reflectance of NIR but increase reflectance of visible wavelengths. The overall result would cause a decrease in the values of these three indices at the grainfilling stage. This is consistent with the observation that LAI commonly peaks at the booting stage and decreases as the growth cycle progresses (Calderini et al., 1997; Aparicio et al., 2000).
Values for the NIR-based indices WI, NWI-1, and NWI-2 decreased from booting to heading, and then increased at grainfilling. These indices are based on reflectance at 970 nm, which is a weak water absorbance band (Peñuelas et al., 1993), and on reflectance at 900 nm and 850 nm, which is caused by multiple reflection and scattering of light in the spongy mesophyll structure (Knipling, 1970). The WI shows an inverse relationship with water status at both the canopy and leaf level (Peñuelas et al., 1993, 1997). Water index assesses the changes in relative water content, leaf water potential, stomatal conductance, and canopy temperature in plants (Peñuelas et al., 1993). Based on the values of WI and the other NIR-related indices, it appears that in general the plant canopy contains a higher total amount of water at heading compared to later growth stages. Assessment of fresh and dry biomass made at these growth stages on the above experiments indicated that there was 14 to 22% more water in the canopy at heading than at grainfilling (data not shown).
Genotype, Growth Stage, and Year Interactions
In this study, we observed significant interaction between growth stages and genotypes in regard to their index values (Table 3). The low association between booting and flowering, and booting and grainfilling for NIR-based indices and PRI (Table 4), indicates that the ranking of genotypes changed between growth stages. On the other hand, RNDVI, GNDVI, and SR showed a higher correlation between booting and flowering and between booting and grainfilling. The interactions of growth stages and indices indicate that care must be taken to identify a suitable growth stage at which the indices will be applied to discriminate most effectively among the genotypes in breeding trials. Aparicio et al. (2002) also reported a significant interaction between growth stages and spectral indices (NDVI and SR) in durum wheat.
Association of Grain Yield with SRI
Several authors (Ball and Konzak, 1993; Peñuelas et al., 1997; Aparicio et al., 2000; Ma et al., 2001; Royo et al., 2003) have reported on the potential use of different spectral indices (mostly in regard to NDVI and SR) to differentiate genotypes for grain yield under diverse environmental conditions. A large proportion of the variation in yield in barley under three salinity levels was explained by spectral indices (Peñuelas et al., 1997), in soybean with three different planting densities and two different soil types under well-watered conditions (Ma et al., 2001), and in durum wheat under different moisture levels and at different locations (Royo et al., 2003). A large part of the variation being explained in these studies may be driven by the diverse environmental conditions to which the crops were exposed. Royo et al. (2003) reported a very high variability in grain yield of different durum wheat genotypes when SRI were combined across environments in a stepwise regression model (nine different moisture levels, nine field experiments, and nine SRI were considered together), but when individual environments were studied on their own, the amount of variation explained decreased considerably. Their study did not give any clear indication of which SRI were suitable for use under different environmental conditions. However, Aparicio et al. (2000) did show a strong relationship between grain yield and spectral indices under rainfed conditions in durum wheat.
This study includes historical landmark genotypes developed by CIMMYT with high yield capacity and with considerable diversity in morphology, plus different sister lines (RLs) expressing a similarly large range of diversity. Strong, consistent phenotypic correlations were observed between the NIR-based indices and grain yield in all experiments in three successive years. Also, high genotypic correlations between grain yield and the NIR-based spectral indices at the heading and grainfilling stages were observed (genotypic correlation values ranged from –0.594 to –0.925 in the above mentioned three different experiments) (data not shown). The estimation of genotypic correlations was done on small genotypic sample sizes in our study and may not be very reliable, but their magnitude should reflect true population values. This indicates that when using the NIR-based indices, the high amount of variation explained is related to variability among genotypes, and not due to any large environmental effects.
The spectral indices based on NIR (WI, NWI-1, and NWI-2) generally showed negative correlations with yield, increasing in value with the advancement of the crop cycle in all experiments (Table 6). These indices all incorporate an indicator of water status in the canopy. With an increasing amount of water in the plant, a decreasing amount of energy at 970 nm is reflected. Hence, the negative sign of the correlations reflect the fact that an increasingly low water status is associated with decreasing yields. Similar results were reported by Peñuelas et al. (1997) and Royo et al. (2003), both under irrigated and water-stressed conditions in durum wheat and barley.
RNDVI, GNDVI, and SR gave significant positive correlations with grain yield at the heading and grainfilling stages in nearly all cases over the 3 yr. Ball and Konzak (1993) and Royo et al. (2003) reported a significant positive correlation between grain yield and RNDVI at grainfilling for spring and durum wheat under well-watered conditions. On the other hand, Aparicio et al. (2000) found a significant correlation between grain yield and RNDVI only at the maturity stage, and not at the other growth stages (booting, heading, anthesis, and milk-grain) in durum wheat under irrigated conditions. We have observed that in most cases RNDVI, GNDVI, and SR showed an increasingly higher correlation with grain yield as growth progressed from booting to flowering or grainfilling (Table 6). Also Ma et al. (2001), when studying soybean, found a clear increasing trend in the correlation values between NDVI and grain yield from full flowering (R2) to seed formation (R5) at three different planting densities in two soil types under well-watered conditions.
All indices studied explained the largest amount of the variation when taken at the heading or grainfilling stages. Our results generally agree with the findings of Ma et al. (2001) in soybean and Royo et al. (2003) in durum wheat under well-watered conditions. In our study it is evident that the mean values over different growth stages gave higher correlations with grain yield, which was not reported by previous authors. The three NIR-based indices measure the water status at the canopy level, while RNDVI, GNDVI, and SR measure the greenness of the canopy, and PRI is an indicator of radiation use efficiency in plants. All the indices are indicative of healthy plant conditions in the field. The repeated measurements on the same genotypes at different growth stages basically accumulate information on the health or condition of the genotypes over a period of time. We hypothesize that the mean values of the indices for the different growth stages represent cumulative information on the health of the canopy, which translates into a higher correlation with final grain yield. This may also be an indication that the underlying genetic correlation is stronger than the phenotypic one.
Since the time required to take the SRI data in the field is just 40 to 45 s plot–1, it would be desirable to take the readings more than once. At a minimum, one measurement at heading and another one at grainfilling might effectively differentiate genotypes for grain yield. Hence, those two stages appear to be the most appropriate time to apply these spectral indices if the objective is to discriminate genotypes for grain yield. Very strong correlations were observed between the overall mean of the SRI and overall grain yield. This strong correlation is also evidence of improved association between SRI and grain yield when more dates were averaged over years and growth stages, which was not reported before.
Our study also demonstrated a high efficiency of the SRI to evaluate the yield performance of genotypes over a period of time, which might be one of the most critical questions for breeders in evaluating a genotype for a particular environment (Table 7). The three NIR-based indices were more successful than others. Our study also showed a very high efficiency for SRI to select superior genotypes for grain yield, and NWI-2 performed better than RNDVI in selecting the top yielding genotypes for grain yield. These results are a definitive indication of the efficiency of NIR-based SRI for selecting superior genotypes for grain yield production.
The performance of the three NIR-based indices was very similar in explaining grain yield variability among genotypes within a year or over a period of time. Normalizing the water index (NWI-1 and NWI-2) did not significantly improve the relationship. Tucker (1979) showed the superiority of a normalized index over a ratio index under water stressed conditions. The normalization partially removed the disturbance caused by external factors such as soil interference, position of sun, illumination, and angle of view. That was not demonstrated in our study. Nonetheless, NWI-2 showed a marginal superiority over the two indices.
Association of SRI with Agronomic Traits
The association between the SRI and yield components was evaluated with the objective of determining if any particular yield component was driving the association with yield. The specific SRI that were best associated with yield (i.e., the NIR-based indices) also showed a greater correlation with these yield components. This observation is consistent with the fact that in all experiments grain yield, grains per square meter, and biomass showed significant positive correlations among themselves (data not presented). Waddington et al. (1986) and Sayre et al. (1997) clearly demonstrated that grain yield is particularly well correlated with grains per square meter in irrigated spring wheat in studies executed in the same location as used in this study. Therefore, it could be hypothesized that the relationship between NIR-based indices and grains per square meter may be the most important basis of the high relationship between NIR-based indices and grain yield.
Future Research Goals
Using an indirect selection tool is appropriate if the genetic correlation between the selected and unselected traits is very high, the heritability is much higher for the selected trait than for the unselected trait, and the correlated response in the unselected trait based on the selected trait is higher than the direct response to selection of the unselected trait (Falconer, 1989). In practice, this combination is rarely obtained. It is also important to consider the time and cost involved in using indirect selection tools compared to the use of grain yield per se as a selection criterion. A research project is currently underway to estimate heritability, expected genetic gain, correlated response to selection for grain yield estimated from the SRI, and the efficiency of selecting superior genotypes for grain yield based on SRI compared to the selection of superior genotype based on yield per se at different geographic locations and at different moisture conditions under the supervision of Oklahoma State University and CIMMTY, Mexico. Efforts are ongoing to develop a new lightweight spectral sensor to take measurements for these NIR-based indices in the field. The new sensor (approximate cost US$ 4000–5000) will facilitate faster measurements in the field compared to the spectroradiometer (cost US$ 30 000) used in the current studies and will bring down the current cost of equipment to a minimum level. These improvements should facilitate the adoption of this selection criterion by breeders and enhance their ability to discriminate genotypes for grain yield in the breeding trials.
|
CONCLUSIONS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Received for publication January 19, 2005.
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. P. Reynolds, C. S. Pierre, A. S.I. Saad, M. Vargas, and A. G. Condon Evaluating Potential Genetic Gains in Wheat Associated with Stress-Adaptive Trait Expression in Elite Genetic Resources under Drought and Heat Stress Crop Sci., December 18, 2007; 47(Supplement_3): S-172 - S-189. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Prasad, B. F. Carver, M. L. Stone, M. A. Babar, W. R. Raun, and A. R. Klatt Genetic Analysis of Indirect Selection for Winter Wheat Grain Yield Using Spectral Reflectance Indices Crop Sci., July 30, 2007; 47(4): 1416 - 1425. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Prasad, B. F. Carver, M. L. Stone, M. A. Babar, W. R. Raun, and A. R. Klatt Potential Use of Spectral Reflectance Indices as a Selection Tool for Grain Yield in Winter Wheat under Great Plains Conditions Crop Sci., July 30, 2007; 47(4): 1426 - 1440. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. A. Babar, M. P. Reynolds, M. van Ginkel, A. R. Klatt, W. R. Raun, and M. L. Stone Spectral Reflectance to Estimate Genetic Variation for In-Season Biomass, Leaf Chlorophyll, and Canopy Temperature in Wheat Crop Sci., March 27, 2006; 46(3): 1046 - 1057. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||
